Improve feature mocking and increase number of features in tests

[w-k-jones]

In many of our tests, we create test features using the existing routines and then test them against hardcoded values. It would be good to improve this mocking so that we can produce both the test features and expected outcomes of feature detection, segmentation or tracking from a single function.

This would also help in adding additional features for existing tests, which could help avoid bugs such as #315 which was not detected as the behaviour works as expected for the first feature detected